First they said it couldn’t be done. Then they said there was no business reason for it to be done. Trouble is, Landen Bain went and did it anyway.
As liaison to health care for the Clinical Data Interchange Standards Consortium (CDISC), Bain has been working for four years to tie together clinical trial systems and the larger health care infrastructure. You won’t meet a calmer, more self-effacing guy. So to get Bain to actually discuss what he’s accomplished is a minor journalistic coup.
Now, in 2009, the vision that Bain and CDISC have long championed has begun to come to pass. In key pilot projects, live trials are grabbing data from real-world electronic health record (EHR) systems. They’re feeding that data into standard electronic data capture (EDC) systems. The ramifications of such efforts are profound. They could extend toward better pharmacovigilance and a higher percentage of trials remaining in the U.S.
“We have two real world implementations going live,” notes Bain. “We have the ability for an EHR to display a pre-populated case report [form] which the EDC system staged in the background.” One key enabling technology is the CDASH effort, which hopes to standardize case report forms. Here’s a ClinPage article on that topic.
Bain is gracious enough to point out that his labor has also benefited from David Iberson-Hurst, another CDISC wizard. Iberson-Hurst has been waging a similarly lonely quest to nudge the industry and its regulatory masters to think about electronic source documents. Which, if you think about, is one way to describe what’s inside an EHR.
Bain’s and Iberson-Hurst’s efforts have always been collaborative. That’s why Bain has no concerns about whether the agency would accept data generated by an EHR. Indeed, the CDISC integration effort will be presented to FDA this spring, as part of a larger discussion around electronic source documents.
What’s in it for the clinical trial community? Bain sees major benefits across the industry. For sites, he notes, there is a certain psychological injury caused when intelligent scientists are asked to do “stupid” things. Such as double data entry. “There is a cost to that,” he says, referring to sites’ frustration and ambivalence about helping industry at all.
For Bain, guessing the cartoon bubble in the head of an investigator—especially in an age of online banking and digital tax filings—does not demand a burst of Einsteinian intellect. Rather, Bain expects, the thought is something like Somebody should be able to figure out how to get the data from there to there without me having to key it in.
Vendors are getting into the mix. When we spoke to Bain this week, he was in the bowels of a Chicago hotel, huddled with 300 or so IT types from the EHR industry. But among the throng were three pioneers from the EDC world: Outcome, Phoenix Data Systems and Nextrials. All were connecting their systems to the EHR universe. “The EDC systems that are forward-looking are looking at this as a new business model,” says Bain. “They want to be in this first.” The two top EDC firms, Bain says, are also considering such efforts.
Providers of EHR systems have a clear motivation. “It offers them a way to help customers do clinical research without them having to develop everything that an EDC system has,” Bain notes.
This reporter had mistakenly anticipated an epic, Darwinian showdown between the EDC and EHR vendor titans. Given how much larger the EHR companies are than the EDC suppliers, it wasn’t hard to imagine the ending of that drama. Except it’s far from how Bain sees the landscape.
He feels EDC companies will survive and cooperate. Bain says that “EDC” is a misnomer, and underemphasizes the work that the EDC vendor community does in converting a complex scientific protocol into a database and research instrument for a regulated industry. Even with the convergence that is on the horizon, Bain thinks there will be plenty of work for both EHR and EDC firms.
And for sponsors? Lilly’s been working with him for a while. Pfizer had been thinking about how to improve spontaneous adverse event reporting. The company heard one of his presentations and recognized that the CDISC technology could be a part of a new project. So Pfizer plugged the CDISC linkage tool into a project with a homegrown EHR in the Boston-based Partners HealthCare network.
Some sponsors see the EHR-EDC convergence as a way to do more than pay lip service to the concerns of the investigator community. “Sites are busy,” says Bain. “Their fundamental reason for existence is patient care. This improves the sponsor’s ability to get new players into the site and investigator pipeline.”
In another live EHR-EDC project in Georgia, Bain spoke to the investigators who had never participated in industry-sponsored research before. The EDC-EHR combination made the work so easy that their view of such projects changed to .... why not participate?
Bain is the first to say that important improvements remain to be done with the EHR-EDC specifications. A variety of governmental and non-governmental organizations (ANSI, HITSP, IHE and, for all we know, others) are continuing the effort to sync up the two data realms. This recent press release from ANSI names 27 organizations that are helping in its effort to ensure a degree of harmony and interoperability in the EHR community. There are big pharmas, computer companies and nonprofits on the list.
But some of these projects are starting to move forward on their own, without Bain’s active involvement. Could capitalism actually work? It appears so. Through the sheer self-interest of the vendors involved, the train for clinical-trial-friendly EHR systems is leaving the station. It’s clear that the EHR-EDC integrations are serious, if embryonic. Some $19 billion in new federal subsidies for EHR systems will only extend the industry’s reach.
Bain is even optimistic that some of the more cantankerous foes of EDC-EHR linkage are starting to see the bigger picture. “The people who have to loosen up a little bit are the biostatisticians,” he says. He jokes that the phase of the moon or the usage of a No. 2 pencil might not necessarily disqualify data for research. Adds Bain: “They need to relax a little bit and not just assume that because data are sourced out of an EHR, that they are somehow suspect.” The tide is starting to turn, he says.